вход по аккаунту


Benefits of polyspecific associations for the GoeldiТs monkey (Callimico goeldii).

код для вставкиСкачать
American Journal of Primatology 54:143–158 (2001)
Benefits of Polyspecific Associations for the Goeldi’s
Monkey (Callimico goeldii)
Interdepartmental Doctoral Program in Anthropological Sciences, State University of New
York at Stony Brook, Stony Brook, New York
Polyspecific associations are an important component of Callimico goeldii
behavior and ecology. On average, Callimico goeldii was found in proximity to or in vocal contact with Saguinus troops (S. fuscicollis and S.
labiatus) during 53% of all time intervals sampled. Polyspecific associations varied considerably between seasons, however, with association rates
peaking during the wet-season month of February (89%) and declining
in the dry season, with the lowest rate (13%) in July. The primary benefits of associations appear to be an increased use of the lower and middle
canopy, and an increase in feeding behaviors during the wet season. Thus,
Callimico goeldii appear to benefit most from associations during the
wet season when fruits are its principal food source. Fruits are eaten
more in the forest canopy than in the understory; thus, an increase in
height use likely permits an increase in the fruit resources on which
Callimico goeldii can forage and feed. In addition, Saguinus groups, with
their smaller home ranges, are likely to be more knowledgeable than
Callimico goeldii about the location and abundance of ripe fruits in their
home ranges. Thus, Callimico goeldii may parasitize Saguinus for their
fruit knowledge by following them through their ranges. In the dry season, limited dietary overlap between Callimico goeldii and Saguinus
groups is likely to make associations less beneficial for Callimico goeldii
as they adopt different foraging and ranging strategies. Am. J. Primatol.
54:143–158, 2001. © 2001 Wiley-Liss, Inc.
Key words: Callimico goeldii; Saguinus fuscicollis; Saguinus labiatus;
mixed species; competition; predation; foraging; home range
Among the unusual characteristics of the neotropical subfamily Callitirichinae,
the marmosets and tamarins, is their tendency to form stable, long-lasting poly-
Contract grant sponsor: Fulbright Scholarship; Contract grant sponsor: National Science Foundation;
Contract grant number: 9815171; Contract grant sponsor: Chicago Zoological Society; Contract grant
sponsor: LSB Leakey Foundation; Contract grant sponsor: Douroucouli Foundation; Contract grant
sponsor: Primate Conservation, Inc.; Contract grant sponsor: Margot Marsh Biodiversity Foundation.
L.M. Porter’s present address is Department of Conservation Biology, Chicago Zoological Society,
Brookfield, IL 60513.
*Correspondence to: Leila Porter, Correo Central, Cobija, Departamento de Pando, Bolivia.
Received 1 March 2000; revision accepted 14 March 2001
© 2001 Wiley-Liss, Inc.
144 / Porter
specific associations [Yoneda, 1981; Pook & Pook, 1982; Terborgh, 1983; Yoneda,
1984; Garber, 1988; Buchanan-Smith, 1990; Heymann, 1990; Norconk, 1990; Peres,
1992a–c; Lopes & Ferrari, 1994; Buchanan-Smith, 1999]. The first studies of wild
Goeldi’s monkeys (Callimico goeldii) showed that they formed polyspecific associations with two tamarin species, the saddle-backed tamarin (Saguinus fuscicollis)
and the red-bellied tamarin (Saguinus labiatus) [Pook & Pook, 1982; BuchananSmith, 1991; Christen & Geissmann, 1994]. The frequency and longevity of these
associations, however, remained uncertain due to the short duration of these investigations and lack of fully habituated animals.
Polyspecific associations are not likely to occur between species with very
disparate behavior. In order to travel and forage together, associated species must
have some degree of ecological similarity which allows coordination of group activities [Terborgh, 1983]. However, with increasing ecological similarity comes a
potential increase in feeding competition. For some species, increasing group size
has been shown to increase feeding competition between individuals within a
single-species group [van Schaik & van Noordwijk, 1986; Janson, 1988; Janson
& Goldsmith, 1995]. Similarly, polyspecific associations increase overall group
size and may increase feeding competition among individuals of a polyspecific
group if dietary overlap is high for spatially restricted resources and/or where
knowledge of resource location is critical to feeding success [Podolosky, 1990;
Terborgh, 1990].
Thus, polyspecific associations are likely to represent a compromise between
competition and compatibility, but the benefits should outweigh any potential
costs incurred through increased feeding competition. Polyspecific associations
are found in a variety of other animals, including ungulates [Fitzgibbon, 1990],
birds [Munn & Terborgh, 1979; Popp, 1988], fish [Landeau & Terborgh, 1986],
and spiders [Hodge & Uets, 1996]. No single factor can universally explain why
these associations occur, as differences between species living in closed and open
habitats, the types of predators present, and resource distribution all appear to
influence association patterns and their formation [Cords, 1990a; Fitzgibbon, 1990;
Terborgh, 1990; Chapman & Chapman, 1996]. This work examines four hypotheses to explain the occurrence of polyspecific associations among Callimico goeldii
and sympatric species of Saguinus.
First, associations may represent chance encounters between different groups
of species sharing a common home range [Waser, 1982, 1984; Whitesides, 1989],
and/or common food resources. Second, polyspecific associations may improve
predator avoidance. Predation is considered to be an important selective force on
the small-bodied callitrichines, which are thought to be prey to a variety of mammalian, avian, and reptilian predators [Peres, 1993]. Although published accounts
of predation on callitrichines are scarce, there is increasing evidence that primates are an important part of the diet of terrestrial and aerial predators [Boinski,
1987; Isbell, 1990; Wright, 1998]. Some studies of African primates support the
antipredation hypothesis, showing that polyspecific associations increase in frequency when predation risks are highest [Oates & Whitesides, 1990; Holenweg
et al., 1995; Nöe & Bshary, 1997].
Increasing group size through polyspecific associations can result in several
potential antipredation benefits. Large polyspecific troops may reduce the probability that any one individual in the troop will be preyed upon, thereby reducing an individual’s risk of predation [Roberts, 1996]. Predator avoidance may
further improve if associated species are active at different heights in the forest,
leading to a greater spread of vigilant individuals in an area than are present in
a monospecific group. With greater vigilance spread the troop can more rapidly
Callimico goeldii Polyspecific Associations / 145
detect predators [Pook & Pook, 1982; Peres, 1989; Peres, 1992b; Caine, 1993;
Peres, 1993; Hardie & Buchanan-Smith, 1997]. Callimico goeldii, Saguinus fuscicollis, and Saguinus labiatus travel at different heights in the forest [Porter,
2000], making it possible for improved detection of predators during associations
through greater vigilance spread. Thus, if polyspecific associations improve predator avoidance, Callimico goeldii is predicted to have increased survival rates while
associated with Saguinus troops. This, however, can only be measured with longterm data on group demographics and predation rates.
It is possible, however, to assess whether associations improve predator avoidance through other measures. The group size effect [Pulliam, 1973] predicts that
individuals in large groups can reduce vigilance behaviors because the group as
a whole can maintain a high vigilance rate. Callitrichines use scanning to remain vigilant to their predators [Caine, 1984; Ferrari & Lopes Ferrari, 1990;
Koenig, 1998], and scanning comprises a large part of the daily activities of
callitrichine species [Ferrari & Lopes Ferrari, 1990; Savage et al., 1996; Koenig,
1998]. Thus, a reduction in scanning can lead to an increase in time spent foraging and feeding if these activities are time limited. The group size effect has
been supported by studies of polyspecific association among some species [Cords,
1990b; Hardie & Buchanan-Smith, 1997] but not for others [Treves, 1998; Garber
& Bicca-Marques, in preparation]. Therefore, if Callimico goeldii benefits from
the group size effect it is predicted that it will decrease scanning and increase
feeding during associations with Saguinus troops.
Third, it has been suggested that some species exploit the resource knowledge of their associates, with one species using another species to lead it to food
resources. Saimiri sciureus follow Cebus apella groups around their territories,
parasitizing the Cebus groups by following them to fruit resources [Terborgh,
1983; Podolosky, 1990]. Cords [1990b] also found that Cercopithecus ascanius uses
Cercopithecus mitus as a guide to food resources in areas where they have comparable diets. If Callimico goeldii is shown to follow at least one Saguinus species, and if Callimico goeldii gain from Saguinus resource knowledge, it is
predicted that Callimico goeldii will consume foods more frequently when associated than when alone.
Fourth, improved insect foraging may result from the flushing of insects from
the middle and upper canopies by one species to an associated species traveling
below [Munn & Terborgh, 1979; Terborgh, 1990]. This has been shown to be one
advantage that Saguinus fuscicollis gains from associations with S. mystax [Peres,
1992b]. If Callimico goeldii benefits from insect flushing from Saguinus, it is
predicted that Callimico goeldii will forage for insects at lower levels than at
least one Saguinus species, and have increased insect feeding rates while associated with Saguinus troops than while it is alone.
Study Site
Research was conducted in northern Bolivia, in the Department of the Pando,
at a field camp at San Sebastian (11° 24′ S, 69° 06′ W, ca. 280 m elevation), 3 km
north of the Rio Tahuamanu and 42 km east of the border of Peru [for a detailed
description see Porter, 2000; in press]. There are 10 species of primates in the
study area, and many species of large predators, including jaguars (Panthera
onca), ocelots (Felis concolor), margays (Felis pardalis), tayras (Eira barbara),
harpy eagles (Harpia harpjya), and tree boas (Corallus hortulanus).
146 / Porter
Study Groups
Following a 7-mo habituation period, observations of one group of Callimico
goeldii were taken each month from April 1998–March 1999. Group scans on the
Callimico goeldii were taken from 7–14 days each month for a total of 957 hr of
observations. Due to the large home range size of the Callimico goeldii group, it
was observed within the home ranges of seven mixed species (Saguinus fuscicollis
and Saguinus labiatus) groups (referred to as groups I–VII). Two of these groups
were observed for this study. Observations of Saguinus fuscicollis in group I were
taken from April 1998–September 1998 (250 observation hr), and in group II from
October 1998–March 1999 (188 observation hr). Observations of Saguinus labiatus
in group I were taken during June 1998–September 1998 (123 observation hr) and
from group II during October 1998–March 1999 (201 observation hr).
Behavioral Methods
Instantaneous group scan samples (10 sec in duration) were collected every
5 min [Martin & Bateson, 1993] from the time the groups left their sleep sites in
the morning until they retired to a sleep site at night. Animals located only after
they moved were not recorded in the scan to avoid biasing the sample towards
more active behaviors. Infants were recorded as a separate individual only when
moving independently of an adult.
For each scan sample the following data were recorded for each individual
observed in the group: its general and specific behavior, and the height class and
habitat type it currently occupied. Behaviors recorded were foraging (including
manual and visual search for food objects), feeding (eating or manipulating a
food object), traveling (including quadrupedal movement, leaping, and vertical
ascent and descent), resting (including scanning, grooming, and sitting) and other
(including play and aggression). Habitat types discussed in this work include: A,
primary forest with open understory (canopy > 15 m, visibility at eye level > 20
m); C, primary forest with dense understory (canopy > 15 m, visibility at eye
level < 20 m); D, bamboo dominant plant species (canopy < 15 m, visibility at eye
level < 20 m); and F, stream’s-edge habitat (canopy > 15 m, visibility < 20 m,
dominant plants ferns and palms). Height classes were defined as follows: 0 = 0
m; 1 = 0–5 m; 2 = 5–10 m; 3 = 10–15 m; 4 = 15–25 m; and 5 = >25 m. The group’s
location (if known) in the trail system was also indicated during each scan.
Interspecific associations, including both physical proximity and vocal contact, were noted using presence–absence sampling between each 5-min scan. Two
species were considered to be in physical association if at least one individual of
another species was within 15 m of the study group under observation. Although
other studies have used greater distance criteria (e.g., 20 m [Cords, 1990b], 25 m
[Nöe & Bshary, 1997], and 50 m [Buchanan-Smith, 1990; Chapman & Chapman,
1996; Wachter et al., 1997]), 15 m was chosen for this study as Callimico goeldii
was found below 5 m during 80% of observations and in dense understory forest
with visibility less than 20 m during 76% of all observations [Porter, 2000]. Thus,
a distance criterion of 15 m ensured that association data were recorded accurately, although more strictly than previous studies.
Two species were considered to be in vocal contact with one another if contact
calls, feeding calls, or alarm calls of another species were audible by the observer
following the focal group. The use of presence–absence sampling allowed for the
recording of vocal associations, behaviors that are rare and short in duration but
Callimico goeldii Polyspecific Associations / 147
very important for group movement and cohesion [Pook & Pook, 1982]. The use of
1:0 sampling should not overestimate association duration [Martin & Bateson, 1993]
as the typical association was generally long-lasting and uninterrupted for much
longer periods of time than the sample interval used (5 min).
Polyspecific group leadership was recorded on 14 days. The leader was defined as the species that physically (for travel or feeding) or temporally (for longcall communication) began an activity that one or more associated species followed.
The initiator of an association was defined as the species that moved towards
another group, with the result that they traveled to within 15 m of one another.
The species initiating the change was noted every time there was a change of
activity. Travel was noted as having begun if one or more individuals of a species
left its resting, feeding, or foraging site and was then followed by the rest of the
group. No leader was noted if there was no species clearly leading the group, or
if an activity change was ambiguous.
Renkonen’s Percentage Similarity Index was used to calculate diet overlap
between species. An index value of 1 indicates complete dietary overlap, and a
value of 0 indicates no dietary overlap [for complete description see Porter, in
press; Wolda, 1981]. For dietary overlap calculations, fruits, nectar, and exudates
were categorized by species, whereas arthropods and fungus were categorized as
food types.
Group scans were conducted rather than focal animal scans, as individuals
of the Callimico goeldii study group were not trapped or marked and therefore
were not easily identifiable. Due to the difficulty locating all individuals of a
group, the behaviors of each individual recorded (independent records) during
one scan were converted to a percentage of the total activity records for that
sample. Weighting scans may incur some error as the study groups were not
observed completely during each scan and did not have similar proportions of all
age classes each month (Table 1). However, as callitrichine groups have been
shown to travel and feed in cohesive units [Garber, 1993; Garber, 2000; Peres,
2000], the behavior of one individual closely approximates the behavior of other
group members during a scan sample. Thus, for this study, each scan had equal
weight in the overall analyses [as in Milton, 1980], and these weighted rates
rather than absolute rates were used throughout analyses.
TABLE I. Group Composition and Monthly Observation Rates
Group composition
Number of individuals
3f, 3m, 1i
3f, 3m, 1i
3f, 3m, 1i
3f, 3m, 1i
2f, 2m, 1j, 1i
2f, 2m, 1j, 1i
2f, 2m, 1j, 1i
2f, 2m, 1j, 1i
2f, 2m, 1j, 1i
1f, 2m
1f, 1m
1f, 1m, 1i
a, adult (unknown sex); f, adult female; m, adult male; i, infant (age 0–6 months); j, juvenile (age 6–12 months).
148 / Porter
Location data recorded each day allowed for an assessment of the area of the
home range Callimico goeldii used for that day. The home range was divided into
eight areas corresponding to the core areas of the sympatric Saguinus fuscicollis
and S. labiatus groups within the Callimico goeldii range. This allowed an assessment of the time Callimico goeldii spent within the range of each Saguinus
group and the time it spent in association with each group.
To determine whether Callimico goeldii associated with Saguinus species more
than expected by chance, I performed the following analysis. I calculated the percentage of observational scans Callimico goeldii occupied 19 (1 ha) plots of forest
within Saguinus fuscicollis group I’s home range, and the percentage of observations that S. fuscicollis group I used these same 19 plots. Using these plot-use
data, I then calculated the probability the two species would occupy the same plot
at the same time (Callimico goeldii % plot use × Saguinus fuscicollis % plot use).
This represents the expected percentage of time the species would be together if
their proximity were merely chance encounters in a commonly used area. The
expected rate was then compared to the actual association rate (based on the much
more conservative criteria of 15-m proximity) using a pairwise t-test (SPSS© 9.0
for Windows) to assess whether associations were due to chance alone.
All percentage data (behavior, diet, and habitat use) were arcsine transformed
before analyses. Using the entire data set, monthly differences in association
rates were compared using a standard factorial ANOVA with unbalanced sample
sizes (SPSS© 9.0 for Windows). A model 1 step-wise multiple regression (SPSS©
9.0 for Windows) was then used to test which of the following variables best
predicted monthly changes in association frequency: group size, dietary overlap,
and frequency of feeding, insectivory, frugivory, and mycophagy.
For comparisons of Callimico goeldii behavior in and out of association,
subsamples of the data were selected from four 1-hr periods (between 8:00–9:00,
10:00–11:00, 12:00–13:00, and 14:00–15:00) across the entire data set to control
for possible temporal variation in behaviors, and hourly averages of behaviors
were used for analyses. Hours during which Callimico goeldii was alone (100% of
all 5-min samples for that hour show no polyspecific associations) were compared to all hours that it was associated (groups were associated with one or
more species during at least 10 out of 12 5-min intervals for that hour). In all
associated samples a 5- or 10-min lapse in association was immediately followed
by at least 1 hour during which the groups were continuously associated; thus
these brief lapses were not indicative of a splitting of the polyspecific troop. Interactions between association status (alone or associated) with season (dry season May–October, or wet season November–April) and time of sample were
examined for all analyses. Unless otherwise stated, interactions among these
fixed factors (time of day, season, or association status) were not statistically
To test for differences in height use during sample periods, the KomolgorovSmirnov test [Sokal & Rohlf, 1995] was used to compare the average height used
during hours associated vs. during hours alone. The median of each height class
was used for calculations (e.g., height class 1 = 2.5 m).
Null Hypothesis
Callimico goeldii was found within Saguinus group I’s home range during
20% of all observations. From these observations, grid location data was avail-
Callimico goeldii Polyspecific Associations / 149
able for 47% of scans (n = 1,224). Location data for Saguinus fuscicollis group I
were available for 30% of scans (n = 1,145). These location data were used as an
estimate of the total frequency that plots were used. Using these frequencies, I
calculated the expected percentage of time the two species would occupy any
given plot at the same time if associations were due to chance (Callimico goeldii
% plot use × Saguinus fuscicollis % plot use) (Table 2). The observed rate was
significantly different than the expected rate (df = 18, t = 5.56, P < 0.001); thus,
associations were not due to chance alone.
Frequency of Associations With Saguinus
Association data were examined for all Callimico goeldii observations taken
throughout the year. Callimico goeldii was found within 15 m of both Saguinus
species at the same time during 22% of observations. Callimico goeldii was found
within 15 m of only Saguinus fuscicollis during 22% of observations, and within
15 m of only S. labiatus during 2% of observations. Callimico goeldii was in
vocal association with a Saguinus group during 7% of intervals. Callimico goeldii
was alone (not within 15 m of, or in vocal contact with either Saguinus species)
47% of the time.
In comparison, data from group II shows that Saguinus fuscicollis was in
association with S. labiatus during 72% of all records (67% within 15 m and 5%
in vocal contact) and with Callimico goeldii during 21% of all records (20% within
15 m and 1% in vocal contact).
Frequency of Associations With Different Saguinus Groups
Callimico goeldii, unlike Saguinus, do not maintain association with only
one group. The home range of the Callimico goeldii study group, with an area of
TABLE II. The Observed and Expected Frequency of Association Between
Callimico goeldii and Saguinus fuscicollis Within Test Plots
Plot Number
Observed Percent
Expected Percent
150 / Porter
150 ha, included the entire home ranges of six Saguinus groups, and parts of two
other groups (Fig. 1). Callimico goeldii spent the most amount of time within the
home range of Saguinus group II, smaller amounts of time with troops I, III, IV,
V, and VI, and only brief periods with troops VII and VIII. The percentage of
time that Callimico goeldii was associated with Saguinus groups in these areas
are indicated in Fig. 1. Associations with a particular Saguinus group (I–VIII)
were maintained for as little as half an hour, and as long as several days.
Patterns of Group Transfers
Three patterns of association transfers were recorded during 40 sequences
when Callimico goeldii moved from one Saguinus territory into another. First,
Callimico goeldii switched directly from one Saguinus troop to another at the
Saguinus troops’ territory boundary (n = 11). Second, Callimico goeldii simply
abandoned their Saguinus associates when it was near a Saguinus territory
boundary. The Callimico goeldii group would then travel by itself in the new
territory, giving long calls intermittently until it encountered a new Saguinus
troop on the same day (n = 11) or after an entire day or more alone (n = 7).
Third, Callimico goeldii would also leave one Saguinus group at their sleep site
in the late afternoon, but then would continue to travel so that in the morning
Callimico goeldii began the day with a different Saguinus troop (n = 11).
Role of Species Within the Polyspecific Group
Callimico goeldii generally initiated associations (66% of records), as it was
usually the first species to start contact calls to locate Saguinus groups. All three
Fig. 1. The home range of one group of Callimico goeldii (dashed line) in relation to the home ranges of
eight Saguinus groups (S. fuscicollis and S. labiatus). The total proportion of time Callimico goeldii spent in
each range (TT) and the proportion of that time spent in association with each Saguinus group (TA) is listed
below each group number. For simplicity only the core areas of a Saguinus troops—not overlapping boundary areas—are shown.
Callimico goeldii Polyspecific Associations / 151
species responded to each other’s contact calls, and this often led the species to
form polyspecific groups immediately after they descended from their sleep trees,
and allowed coordination of group activities throughout the day. Once the
polyspecific group had formed, Saguinus labiatus almost always led group travel
and feeding at common food resources, and it was generally the species that
initiated resting bouts (Fig. 2). Neither Callimico goeldii nor Saguinus fuscicollis
were observed directing any aggression (charging or chasing) at S. labiatus.
Callimico goeldii and Saguinus fuscicollis engaged in aggressive interactions on
39% of observation days, but as these episodes were brief; they formed less than
1% of observations during their association. Thus, feeding competition by aggression appeared to be minimal among species.
Effects of Polyspecific Association on Callimico goeldii Behavior
Habitat use by Callimico goeldii varied with association status and by season. The mean frequency with which stream edge habitats were occupied by
Callimico goeldii was higher while alone than while associated (F[1, 180] = 10.64, P
< 0.001) and more in the dry season than the wet season (F[1, 180] = 3.92, P < 0.05)
[dry season (alone 12%, associated 3%), wet season (alone 5%, associated 0%)
(Fig. 3a)]). In contrast, use of primary forest with open understory was highest
during the wet season during associations (wet season [associated 8%, alone 0%],
dry season [associated 2%, alone 0%], F[3, 178] = 5.58, P < 0.05 (Fig. 3b)).
The mean frequency with which Callimico goeldii used bamboo forest (habitat D) was higher when alone (16%) than when in association with Saguinus
troops (3%) regardless of season (F[1, 180] = 15.10, P < 0.001). In contrast, the mean
frequency with which Callimico goeldii used habitat C (primary forest with dense
understory) was higher while Callimico goeldii was associated (88%) than when
it was alone (67%) regardless of season (F[1, 180] = 18.68, P < 0.001).
Fig. 2. Species initiating mixed group activities for different activities. Cg = Callimico goeldii, Sl = Saguinus
labiatus, and Sf = Saguinus fuscicollis.
152 / Porter
Fig. 3. Frequency with which Callimico goeldii behavior
changed according to association status and season: (a) streamedge habitat use, (b) primary forest with open understory use,
and (c) feeding. Error bars show the SEM ± 1.0 SD.
Height use varied with association status in habitat C. By using data only
from habitat C, I controlled for potential differences in the height at which an
animal travels in habitats with different vegetation profiles. Callimico goeldii
was found at higher heights (average 3.96 m) while associated than while alone
(average 3.14 m) within habitat C (Z = 1.43, P < 0.05).
I also examined scanning within habitat C. If data from all height classes
are combined, scanning is significantly less when Callimico goeldii is part of a
polyspecific group (mean 53%) than when it is alone (mean 62%) (F[1, 98] = 10.77,
P < 0.001). If, however, data from only the same height class is used (height class
1: 0–5 m) to control for potential differences in scanning rates in different height
classes, there is no significant difference in scanning, although on average
Callimico goeldii scanned less during associations (57%) than when alone (65%)
(F[1, 29] = 2.26, P < 0.14). Thus, Callimico goeldii remains highly vigilant through
scanning within height class 1 regardless of association status.
Although not significant, there was a trend toward increased frequency of
travel by Callimico goeldii when it was in association (14% of observations) vs.
when it was alone (11% of observations) (F[1, 98] = 3.46, P < 0.07). However, as
distances traveled were not taken, it is not certain if increased travel frequency
corresponds with increased day ranges. Increased travel may be indicative of
increased feeding competition within polyspecific troops. Therefore, the effects of
associations on feeding behaviors were examined.
Feeding frequencies were compared between when Callimico goeldii was alone
vs. when it was associated using samples collected only within habitat C. In this
way the analysis controlled for potential differences in food availability between
habitats. Feeding rates were significantly higher during the wet season while
Callimico goeldii was associated (13%) than when it was alone (2%), whereas
feeding rates varied little in the dry season between times when it was associ-
Callimico goeldii Polyspecific Associations / 153
ated (7%) and when it was alone (8%) (F[3, 96] = 13.31, association × season P <
0.001 (Fig. 3c)).
I examined the types of food eaten during samples in habitat C (in which
feeding occurred) to determine if associations affected the frequency with which
different types of food were consumed. First, I examined whether arthropod foraging was affected by flushing of arthropods by one species to another. Using the
height at which arthropods were eaten as an approximation of the height at
which they were captured, I found that Saguinus labiatus captured insects at
higher heights than the other two species (mean insect eating heights: Callimico
goeldii 2.94 m (92% of insect feeding records were below 5m); Saguinus fuscicollis
3.88 m; S. labiatus 11.78 m). Therefore, as Saguinus labiatus is the species that
potentially flushes insects to Callimico goeldii, I compared arthropod feeding only
during hours when Callimico goeldii was in association with both Saguinus
labiatus and Saguinus fuscicollis to hours in which Callimico goeldii was alone.
The proportion of feeding time devoted to insects did not differ significantly between hours associated vs. hours alone (F[1, 63] = 1.38, P < 0.25). Thus although
Saguinus labiatus potentially flushes insects from the lower and middle canopies to Callimico goeldii in the understory, its proximity does not result in significant differences in the frequency of insectivory.
In addition, there were no significant increases in the frequency of frugivory
or mycophagy when Callimico goeldii was associated as compared to when it was
alone. Thus, associations appear to result in a general increase in feeding behaviors, not an increase in consumption of any particular food type. There were
seasonal differences, however, in Callimico goeldii feeding behaviors, with
frugivory higher in the wet season (43%) than the dry season (19%) (F[1, 76] = 7.48,
P < 0.01) and mycophagy higher in the dry season (25%) than the wet season
(9%) (F[1, 76] = 4.73, P < 0.05).
Although there were no significant changes in the frequency with which certain foods were consumed during polyspecific associations, feeding behaviors appear to be closely linked to polyspecific association rates. I therefore tested,
through multiple regression analyses, whether changes in feeding behaviors during different months were good predictors of association rates. I used monthly
dietary overlap (with Saguinus fuscicollis or S. labiatus), and the percentage of
food types in feeding records (fruits, fungus, or insects), for the multiple regression analyses. Analyses showed that the percentage of feeding records that were
on fruits (frugivory) was the best predictor of monthly association rates (F[5, 11] =
6.08, P < 0.05, P ≤ 0.05 to add, P ≥ 0.10 to remove (Fig. 4a)). It can also be seen
that although frugivory was the best predictor, mycophagy was roughly inversely
proportional to association rates (Fig. 4a), and monthly dietary overlap with both
Saguinus species roughly parallel monthly association rates (Fig. 4b).
Groups of Saguinus fuscicollis and S. labiatus in this study, as in previous
studies [Buchanan-Smith, 1990; Peres, 1992c; Buchanan-Smith, 1999], maintain
associations exclusively with one group of the other species, sharing common
home ranges, and defending common territories. In contrast, Callimico goeldii
have a home range approximately six times larger than Saguinus, and move
between different Saguinus groups throughout their range (Fig. 1). This is similar to association patterns described for Saimiri with Cebus apella [Terborgh,
1983; Podolosky, 1990], with one Saimiri group associating with multiple Cebus
groups throughout its large home range. Associations were generally initiated by
154 / Porter
Fig. 4. The mean percentage of time Callimico goeldii was in association with Saguinus troops, plotted
with (a) the frequency of frugivory and mycophagy, and (b) dietary overlap values with Saguinus fuscicollis
and S. labiatus.
Callimico goeldii through contact calls that were responded to by both Saguinus
species. Once together, however, Saguinus labiatus led group activities, such as
travel and feeding, and initiated resting bouts.
Associations between Callimico goeldii and Saguinus troops occurred much
more frequently than expected by chance. On average, Callimico goeldii was found
in association with Saguinus groups during 53% of observations. Monthly association rates varied, however, from 89% of observations in the wet-season month
of February to 13% in the dry-season month of July. Multiple regression analyses showed that association frequency was best predicted by the frequency of
fruit feeding: months of high frugivory were also the months of high association
Callimico goeldii Polyspecific Associations / 155
rates (Fig. 4a). In addition, dietary overlap rates roughly parallel association
rates. This suggests that polyspecific associations are linked closely with foraging compatibility and foraging benefits.
Foraging compatibility has been proposed to constrain polyspecific associations among other primate species. Cords [1990a] found that Cercopithecus species associated more frequently at a site where they had higher dietary overlap
than the site where dietary overlap was low. Furthermore, Chapman and
Chapman [2000] found that food availability and interspecific feeding competition limited associations between red colobus groups and groups of other primate
species. The data from the present study suggest that differences in foraging and
feeding strategies among Callimico goeldii, Saguinus fuscicollis, and S. labiatus
when fruits are scarce reduces their compatibility with one another.
During fruit scarcity, Saguinus species eat nectar and exudates while
Callimico goeldii eats fungi [Porter, in press]. Fungi are eaten by Callimico goeldii
throughout the year, particularly in the dry season, but are rarely eaten by
Saguinus species [Porter, in press]. Fungi are patchy and ephemeral resources
that are found more often in stream-edge and bamboo habitats [Hanson, 2000],
habitats that appear to be less useful for Saguinus troops. Indeed, Callimico
goeldii habitat use varied significantly with association status and season.
Callimico goeldii used both stream-edge and bamboo habitats more while alone
than while associated, and used stream-edge habitats more in the dry season,
when mycophagy was highest. These patterns of habitat use suggest that foraging for fungi limits Callimico goeldii associations with Saguinus species that
forage on other resources (exudates and nectar) found in other habitats [Porter,
in press].
Callimico goeldii was found on average nearly 1 m higher during associations than alone. This increase in height use appears to allow Callimico goeldii
to expand the area in which it forages and feeds, a benefit that is particularly
important during the wet season, when feeding rates increased during polyspecific
associations. During the wet season, fruits were the principle food resource for
Callimico goeldii [Porter, 2000; in press], and 89% of these fruits were consumed
above 5 m from the ground [Porter, 2000; in press)]. Thus it is during months
when fruits are available that Callimico goeldii would benefit most from increasing height use.
Although many studies have reported decreased scanning rates during associations, Callimico goeldii showed no reduction in scanning behavior if habitat
type and height were controlled for in the analysis. Thus there is no support for
the group effect: associations do not permit a decrease in scanning rates. However, overall predation risk is likely to be lower in an associated group due to
decreased probability of capture [Roberts, 1996] and increased number and spread
of vigilant individuals in a larger group [Pook & Pook, 1982; Peres, 1989; Peres,
1992b; Caine, 1993; Peres, 1993; Hardie & Buchanan-Smith, 1997]. Increased
predator avoidance in the mixed group is likely to explain why Callimico goeldii
leaves its preferred habitat (dense understory [Porter, 2000]) for more exposed
habitats (primary forest and the lower and middle canopies) more frequently
while in the presence of Saguinus groups. Indeed, this is supported by anecdotal
observations that if Callimico goeldii arrived at a fruiting tree before nearby
Saguinus groups, it would wait under the tree until Saguinus arrived before
climbing into the canopy to feed (Porter and Hanson, personal observations).
In addition, as Callimico goeldii follows Saguinus labiatus while in an associated group, it is possible that C. goeldii uses S. labiatus as a guide to fruit
resources in the canopy, thereby increasing the height at which it forages and
156 / Porter
feeds. Parasitic relationships, in which one species increases food consumption
by following a species with greater fruit resource knowledge, have been demonstrated with captive tamarins [Prescott & Buchanan-Smith, 1999], with Saimiri
sciureus and Cebus apella [Terborgh, 1983; Podolosky, 1990] and in associated
Cercopithecus groups [Cords, 1990a]. In this study, although associations led to
an increase in feeding during the wet season, they did not lead to an increase in
frugivory (as percentage of diet). Thus, while Callimico goeldii may eat fruits
regardless of association status, it can eat fruits more frequently when in an
associated group.
While associations improve fruit feeding and foraging, Callimico goeldii does
not appear to gain from associations with Saguinus through improved insectivory.
Studies show that orthopterans, the principal type of arthropods eaten by
Callimico goeldii, Saguinus fuscicollis, and S. labiatus (41%, 58%, and 100% of
insect records, respectively [Porter, in press]), are at very low abundance and
sparsely distributed in tropical forests [Penny & Arias, 1982; Porter, in press].
Thus, Saguinus troops are not likely not to lead Callimico goeldii to orthopteran
resources. Furthermore, although Saguinus labiatus could potentially flush insects to Callimico goeldii, given the differences in their foraging heights, there
were no significant increases in insectivory for Callimico goeldii during polyspecific
associations. Finally, traveling higher in the forest would not aid Callimico goeldii
insect foraging, as insects were eaten almost exclusively in the understory.
Improved fruit foraging, therefore, appears to be the principal benefit that
Callimico goeldii gains from polyspecific associations. Given recent evidence suggesting that polyspecific associations can vary considerably over small spatial
and temporal scales [Chapman & Chapman, 2000], it is important that further
studies be conducted in order to compare association patterns from this study
with those of Callimico goeldii in other areas. It is particularly important to
more closely assess the role that Callimico goeldii, Saguinus labiatus, and S.
fuscicollis have in predator detection in the understory and canopy. The high
rates of scanning by Callimico goeldii, regardless of its association status, suggest that one reason Saguinus groups may actively maintain associations with
Callimico goeldii is to increase their ability to avoid understory predators.
Thanks to the Ministerio de Desarollo y Medio Ambiente of La Paz, the
Colección Boliviana de Fauna, and the Herbario Nacional de Bolivia for assistance
in obtaining my research permits. Special thanks to my guide, Edilio Nacimento
B., and my research assistants, Kristin Donaldson, Laura Johnson, and Gonzalo
Calderon V., for all their help in data collecting. Thanks to Drs. Patricia Wright,
John Fleagle, Charles Janson, Diane Doran, and Anita Christen for their help in
formulating this project and making comments on this manuscript.
Boinski S. 1987. Birth synchrony in squirrel monkeys (Saimiri oerstedi). Behav Ecol
Sociobiol 21:393–400.
Buchanan-Smith H. 1990. Polyspecific association of two tamarin species, Saguinus
labiatus and Saguinus fuscicollis, in Bolivia. Am J Primatol 22:205–214.
Buchanan-Smith H. 1991. Field observations
of Goeldi’s monkey, Callimico goeldii, in
northern Bolivia. Folia Primatol 57:102–105.
Buchanan-Smith H. 1999. Tamarin polyspecific associations: forest utilization and
stability of mixed-species groups. Primates
Caine NG. 1984. Visual scanning by tamarins. Folia Primatol 43:59–67.
Callimico goeldii Polyspecific Associations / 157
Caine N. 1993. Flexibility and co-operation
as unifying themes in Saguinus social organization and behaviour: the role of predation pressure. In: Rylands A, editor.
Marmosets and tamarins. New York: Oxford University Press. p 200–219.
Chapman CA, Chapman CA, Chapman LJ.
1996. Mixed-species primate groups in the
Kibale Forest: Ecological constraints on
association. Int J Primatol 17:31–50.
Chapman C, Chapman L. 2000. Interdemic
variation in mixed-species association patterns: common diurnal primates of Kibale
National Park, Uganda. Behav Ecol Sociobiol 47:129–139.
Christen A, Geissmann T. 1994. A primate
survey in northern Bolivia, with special
reference to Goeldi’s monkey, Callimico
goeldii. Int J Primatol 15:239–273.
Cords M. 1990a. Mixed-species association
of East African guenons: general patterns
or specific examples? Am J Primatol 21:
Cords M. 1990b. Vigilance and mixed-species
association of some East African forest monkeys. Behav Ecol Sociobiol 26:297–300.
Ferrari S, Lopes Ferrari M. 1990. Predator
avoidance behavior in the buffy-headed
marmoset, Callithrix flaviceps. Primates
Fitzgibbon CD. 1990. Mixed-species grouping in Thomson’s and Grant’s gazelles: the
antipredator benefits. Anim Behav 31:
Garber PA. 1988. Diet, foraging patterns,
and resource defense in a mixed species
troop of Saguinus mystax and Saguinus
fuscicollis in Amazonian Peru. Behaviour
Garber P. 1993. Feeding ecology and behavior of the genus Saguinus. In: Rylands AB,
editor. Marmosets and tamarins. Oxford:
Oxford University Press. p 273–295.
Garber P. 2000. The behavioral ecology of
mixed species troops of Callimico goeldii,
Saguinus labiatus and S. fuscicollis in
northwestern Brazil. Am J Phys Anthropol
Suppl 30:155.
Garber P, Bicca-Marques J. In preparation.
Evidence of predator sensitive foraging in
small and large-scale space in free-ranging tamarins (Saguinus fuscicollis, Saguinus imperator and Saguinus mystax).
In: Miller L, editor. Eat or be eaten: predator sensitive foraging in primates. Cambridge: Cambridge University Press.
Hanson A. 2000. Habitat use in relation to
diet, with particular emphasis on mycophagy, by Callimico goeldii in Pando,
Bolivia. Masters thesis, State University
of New York at Stony Brook.
Hardie S, Buchanan-Smith H. 1997. Vigilance
in single- and mixed-species groups of tama-
rins (Saguinus labiatus and Saguinus fuscicollis). Int J Primatol 18:217–234.
Heymann EW. 1990. Interspecific relations
in a mixed-species troop of moustached
tamarins, Saguinus mystax, and saddleback tamarins, Saguinus fuscicollis (Platyrrhini: Callitrichidae), at the Rio Blanco,
Peruvian Amazonia. Am J Primatol 21:
Hodge MA, Uets GW. 1996. Foraging advantages of mixed-species association between
solitary and colonial orb-weaving spiders.
Oecologia 107:578–587.
Holenweg A-K, Nöe R, Schabel M. 1995.
Waser’s gas model applied to associations
between red colobus and diana monkeys
in the Tai National Park, Ivory Coast. Folia Primatol 67:125–136.
Isbell L. 1990. Sudden short-term increase
in mortality of vervet monkeys (Cercopithecus aethiops) due to predation in Amboseli
National Park, Kenya. Am J Primatol
Janson CH. 1988. Intra-specific food competition and primate social structure: a
synthesis. Behaviour 105:1–17.
Janson CH, Goldsmith ML. 1995. Predicting
group size in primates: foraging costs and
predation risks. Behav Ecol 6:326–336.
Koenig A. 1998. Visual scanning by common
marmosets (Callithrix jacchus): functional
aspects and the special role of adult males.
Primates 39:85–90.
Landeau L, Terborgh J. 1986. Oddity and
the “confusion effect” in predation. Anim
Behavior 34:1372–1380.
Lopes M, Ferrari S. 1994. Foraging behavior of a tamarin group (Saguinus fuscicollis weddeli) and interactions with
marmosets (Callithrix emiliae). Int J
Primatol 15:373–387.
Martin P, Bateson P. 1993. Measuring behaviour: an introductory guide. Cambridge: Cambridge University Press.
Milton K. 1980. The foraging strategy of
howler monkeys: a study in primate economics. New York: Columbia University
Munn CA, Terborgh JW. 1979. Multi-species
territoriality in neotropical foraging flocks.
Condor 81:338–347.
Nöe R, Bshary R. 1997. The formation of
red colobus-diana monkey associations under predation from chimpanzees. Proc R
Soc Lond B 264:253–259.
Norconk M. 1990. Mechanisms promoting
stability in mixed Saguinus mystax and
S. fuscicollis troops. Am J Primatol 21:
Oates JF, Whitesides GH. 1990. Association
between olive colobus (Procolobus verus),
diana guenons (Cercopithecus diana), and
158 / Porter
other forest monkeys in Sierra Leone. Am
J Primatol 21:129–146.
Penny N, Arias J. 1982. Insects of an Amazon forest. New York: Columbia University Press.
Peres C. 1989. Costs and benefits of territorial defense in wild golden tamarins,
Leontopithecus rosalia. Behav Ecol Sociobiol 25:227–233.
Peres C. 1992a. Consequences of joint-territoriality in a mixed-species group of
tamarin monkeys. Behavior 123:220–246.
Peres CA. 1992b. Consequences of joint-territoriality in a mixed-species group of
tamarin monkeys. Behavior 123:220–246.
Peres CA. 1992c. Prey-capture benefits in a
mixed-species group of Amazonian tamarins, Saguinus fuscicollis and S. mystax.
Behav Ecol Sociobiol 31:339–347.
Peres CA. 1993. Anti-predation benefits in
a mixed-species group of Amazonian tamarins. Folia Primatol 61:61–76.
Peres C. 2000. Territorial defense and the
ecology of group movements in small-bodied neotropical primates. In: Boinski S,
Garber P, editors. On the move: how and
why animals travel in groups. Chicago:
Chicago University Press. p 100–124.
Podolosky RD. 1990. Effects of mixed-species association on resource use by Saimiri
sciureus and Cebus apella. Am J Primatol
Pook A, Pook G. 1982. Polyspecific association between Saguinus fuscicollis, Saguinus labiatus, Callimico goeldii and
other primates in north-western Bolivia.
Folia Primatol 38:196–216.
Popp JW. 1988. Scanning behavior of finches
in mixed-species groups. Condor 90:510–511.
Porter L. 2000. Behavior and ecology of the
Goeldi’s monkey (Callimico goeldii) in
northern Bolivia. PhD thesis, State University of New York at Stony Brook.
Porter L. In press. Dietary differences
among sympatric callitrichines in northern Bolivia: Callimico goeldii, Saguinus
fuscicollis and Saguinus labiatus. Int J
Prescott MJ, Buchanan-Smith HM. 1999.
Foraging efficiency in single- and mixedspecies troops of tamarins. Folia Primatol
Pulliam H. 1973. On the advantages of
flocking. J Theor Biol 38:419–422.
Roberts G. 1996. Why individual vigilance
declines as group size increases. Anim
Behav 51:1077–1086.
Savage A, Snowdon C, Giraldo L, Soto L.
1996. Parental care patterns and vigilance
in wild cotton-top tamarins (Saguinus oedipus). In: Norconk M, Rosenberger A,
Garber P, editors. Adaptive radiations of
neotropical primates. New York: Plenum
Sokal RR, Rohlf FJ. 1995. Biometry, 3rd edition. New York: WH Freeman. 887 p.
Terborgh J. 1983. Five New World primates.
Princeton: Princeton University Press.
Terborgh J. 1990. Mixed flocks and polyspecific associations: costs and benefits of
mixed groups to birds and monkeys. Am J
Primatol 21:87–100.
Treves A. 1998. The influence of group size
and neighbors on vigilance in two species of arboreal monkeys. Behaviour 135:
van Schaik CP, van Noordwijk M. 1986. The
hidden costs of sociality: intra-group variation in feeding strategies in Sumatran
long-tailed macaques (Macaca fascicularis). Behaviour 99:296–315.
Wachter B, Schabel M, Nöe R. 1997. Diet
overlap and poly-specific associations of
red colobus and diana monkeys in the Tai
National Park, Ivory Coast. Ethology
Waser P. 1982. Primate polyspecific associations: do they occur by chance? Anim
Behav 30:1–8.
Waser PM. 1984. “Chance” and mixed-species associations. Behav Ecol Sociobiol
Whitesides GH. 1989. Interspecific associations of diana monkeys, Cercopithecus
diana, in Sierra Leone, West Africa: biological significance or chance? Animal
Behav 37:760–776.
Wolda H. 1981. Similarity indices, sample
sizes and diversity. Oecologia 50:296–302.
Wright P. 1998. Impact of predation risk on
the behavior of Propithecus diadema edwardsi in the rain forest of Madagascar.
Behaviour 135:483–512.
Yoneda M. 1981. Ecological studies of Saguinus fuscicollis and Saguinus labiatus
with reference to habitat segregation and
height preference: Overseas Research Reports of New World monkeys. Primate Research Institute, Kyoto University, Kyoto,
Japan. p 43–50.
Yoneda M. 1984. Comparative studies on
vertical separation, foraging behavior and
traveling mode of saddle-backed tamarins
(Saguinus fuscicollis) and red-chested
moustached tamarins (Saguinus labiatus)
in northern Bolivia. Primates 25:414–422.
Без категории
Размер файла
237 Кб
callimico, goeldiтs, polyspecific, goeldii, associations, monkey, benefits
Пожаловаться на содержимое документа